AN ABSTRACT OF THE THESIS OF
KATHLEEN STEELE CORUM for the degree MASTER OF SCIENCE (Name of student) (Degree) in ANIMAL BREEDING AND GENETICS presented on/ (Major department) (Date
Title: GENOTYPE X ENVIRONMENT INTERACTIONS IN SHEEP
Abstract approved: Redacted for privacy (Signature) William D. Hohenboken
The 865 ewe production records taken from three lamb crops, were analyzed to study breed effects, heterosis, environmental effects and interactions between them, and breed and heterosis x environment inter- action effects on ewe production traits. In each of two environments, approximately 144 ewes were mated in a three breed reciprocal cross design each year. The resulting 757 parturitions produced 1263 lambs.
Hampshire, Suffolk and Willamette sheep from university flocks were used.
Two rams of each breed were used in each environment each year. One location was rolling hill land, the other was level, irrigated, valley bottom land, both near Corvallis, Oregon.
Least squares analyses of variance were computed on ewe production traits which included fertility, lambs born, percent survival to weaning,
lambs weaned and pounds of lamb weaned. None of the effects tested in
the analysis of fertility showed significance.On the other traits, year was significant except for survival. Location was significant for pounds weaned per ewe mated and for lambs born, percent weaned and pounds weaned
per ewe lambing. The main effect of the hill land was superior except
for lambs born per ewe which lambed. Dam age was significant except for
survival. The age effect on fertility was not tested. Otherwise pro- duction increased with age. Dam weight change during the mating season was significant for pounds weaned per ewe mated and forprolificacy.
Pounds weaned increased with weight gain while prolificacy decreased.
Lambing date was significant for lambs born and lambs weaned per ewe
lambing. Prolificacy increased as the lambing season progressed.
Location x year interaction effects on pounds of lamb weaned per ewe
mated and on lambs weaned, percent weaned and pounds of lamb weaned per
ewe which lambed were significant.
Dam breed and dam x location interaction were significant for
pounds weaned per ewe mated. The overall and hill land ranking of ewe
breeds was Willamette, Suffolk, then Hampshire. The interaction re.
sulted primarily from a disproportionate increase in the performanceof
the Willamette ewes on the hill land. The Suffolk ewes were superior on
the irrigated pasture. This suggests that a superior adaptation of
Willamette ewes to hill land has resulted during their developmentand
selection there. Sire x year was significant for lambs born. The inter-
action effect alone, however, in no case was more than 0.10lamb so its
importance is questionable. Heterosis and heterosis x location inter-
action were significant for pounds weaned per ewe mated. Heterosis was
31.3% for pounds weaned per ewe mated on the valley pastureand 8.2% on
the hill pasture. These results suggest that heterosis is greaterwhen
conditions for the expression of a given trait aresuboptimal. Genotype x Environment Interactions in Sheep
by
Kathleen Steele Corum
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the degree of
Master of Science
Commencement June 1974 APPROVED: Redacted for privacy Assistant Professor of Animal Breeding and Genetics in charge of major Redacted for privacy Head of Department of Animal Science Redacted for privacy
Dean of Graduate School
Date thesis is presented
Typed by Linda Peoples for Kathleen Steele Corum Acknowledgements
The author expresses appreciation to Drs. W. D. Hohenboken, K. Rowe and L. Swanson for guidance and counsel in the preparation of the study program and of this thesis.
Thanks are extended to Drs. Ralph Bogart and C. W. Fox for the de- sign of this experiment, and again with Dr. W. D. Hohenboken for its administration. Appreciation is shown to Mr. Millard Shelton and
Mr. Lloyd Wescott for collection of the data. Thanks are extended also to Dr. Kenneth Rowe for advice and assistance in the computer analysis of this data.
A special acknowledgement is made to my husband, Kell, for his encouragement, support and good humor during the thesis preparation. TABLE OF CONTENTS
I. Introduction Objectives
II. Review of Literature 4
III. Materials 11 Source of Experimental Animals 11 Mating Scheme 11 Environments 12 Management Practices 14
IV. Methods 16 Components of the Fertility Model 18 Components of the Model Analyzed on a Per Ewe Mated Basis....19 Components of the Model Analyzed on a Per Ewe Lambing Basis..20
V. Results and Discussion 22 Environmental Effects and Interactions 26 Breed Effects and Breed x Environment Interactions 35 Heterosis and Heterosis x Environment Interaction 42
VI. Summary and Conclusions 47
VII. Bibliography 51
VIII. Appendices A. Breed Means for Fertility 55 B. Breed Means for Ewe Production Traits 56 C. Breed/Year/Location Means 58 D. Location Means 68 69 E. Heterosis LIST OF ILLUSTRATIONS
Figure Page
1 Location x Year Interaction for Lambs Weaned per Ewe 31 Lambing
2 Location x Year Interaction for Percent Weaned 32
3 Location x Year Interaction for Pounds Weaned per 33 Ewe Mated
Location x Year Interaction for Pounds Weaned per 34 Ewe Lambing
Breed Effects on Pounds Weaned per Ewe Mated 37
Dam x Location Interaction for Pounds Weaned per 39 Ewe Mated
7 Sire x Year Interaction for Lambs Born per EweMated 40
8 Sire x Year Interaction for Lambs Born per Ewe Lambing 41
Means of Specific Crosses for Pounds Weaned per Ewe 43 Mated 44 10 Heterosis x Location Interaction LIST OF TABLES
Table Page
I. Mating SchemeUsed Each Year 13
II. Least SquaresAnalysis of Variance forEwe Fertility 23
III. Least SquaresAnalyses of Variance forEwe Productivity 24 Based on EwesMated (M)
IV. Least SquaresAnalyses of Variance forEwe Productivity 25 Based on EwesWhich Lambed (L)
V. EnvironmentalEffects on Ewe Production Traits 27
VI. Breed Effects on Ewe Production Traits 36 GENOTYPE X ENVIRONMENT INTERACTIONS IN SHEEP
I. INTRODUCTION
Increases in the costs of sheep production, particularly due to rising grain prices, are resulting from increasing human consumption of grain products and the competition for grains from more efficient domes- tic species. This and the decreasing wool returns resulting from compe-
tition of synthetic fibers with wool are placing increased pressure on
the sheep industry to improve the efficiency of production and to put more emphasis on lamb production. Utilization of less costly by-product
feeds and range lands and increasing the production of lambs are vital
to the prosperity of the sheep industry. Total pounds of lamb weaned per
ewe mated is the best single measure of a flock's production, consequent-
ly more attention is being given to the number of lambs produced per ewe
as well as to the quality of the individual lambs.
There is a large amount of evidence which indicates that breed per-
formance is influenced by environment in a non-additive manner. An
obvious example is the suitability of Merino and Rambouillet sheep to
arid climates and of Suffolk and Hampshire sheep to areas of higher
rainfall. The arid climate may not provide the forage necessary for the
larger, faster growing Down breeds to attain their genetic potential for
growth. In the arid climate, the better suited wool breeds may actually
excel in the characteristic, growth, for which the Down breeds were
developed. The extent and importance of genotype x environment inter-
actions needs to be determined.The relative economic suitabilities of feeds and range lands for sheep can change. With these changes, the relative merits of sheep breeds may also change.
There are many breeds of sheep throughout the world. The British, for example, have developed a large number of breeds, each with a specific purpose or local adaptation. The Scots often say that the
Cheviot is best adapted to grass ranges, the Blackface to heather covered hills. "However, the reported belief among shepherds in the North of
England (Clarke, 1963) that the Herdwick should be grazed 'on the wet side of a fell, the Swaledale on the dry side' may be carrying the idea of local adaptation a bit far." (Carter et al., 1971) Some of this
great multiplicity of breeds may be attributable to personal preference and to poor roads and transportation, but some tangible differences have
developed to justify the discrimination between breeds. How many breeds
are needed though? Singh et al. (1967) points out that genetic improve- ment of breeds is costly and time consuming and suggests that limited
facilities be devoted to fewer, more promising breeds and that those with lower performance be discarded. The question suggested here is
whether breeds can be so unequivocally ranked.
The existence of genotype x environment interactions which involve
a change in rank of genotypes in different environments would make the
estimation of overall breed effects meaningless. The existence of
economically important genotype x environment interactions could render
a new dimension to genetic improvement, performance tests and selection
programs. Breeds and/or mating systems would have to be selected,
tested and developed in the environment of their anticipated use with 3
the expectation of poorer performance elsewhere, or alternativelybreeds and/or mating systems would have to be selected for adaptabilityunder a wide range of environments and maximumsuitability in any of the environments sacrificed.
Objectives
The objectives of this experiment were to evaluate theHampshire,
Suffolk and Willamette breeds of sheep and to documenttheir strengths and weaknesses; to determine whether heterosis existedin crosses between these three breeds; to examine environmental effects andthe interactions between them; and to examine breed and heterosis interactionswith en- vironment in ewe production traits. The models included sire breed,
dam breed, location, year, sires nested within locationswithin years,
dam age, dam weight change during the mating season,lambing date, and
the two-, and three-way interactions between location, year,sire breed
and dam breed. Ewe production traits analyzed were: fertility, lambs
born, lambs weaned, and pounds weaned on the basisof all ewes mated,
and the latter three plus percent weaned on thebasis of those ewes
which actually lambed. 4
II. REVIEW OF LITERATURE
Interest in the effect of environment on selection effectiveness perhaps began with John Hammond who wrote on the subject as early as
1936. In 1941, Hammond identified the importance of environment as a limiting factor in the expression of hereditary potential. He recommended that environment be manipulated to stimulate the most desirable expres- sion of traits so that the ultimate hereditary potential of the individ- uals could be identified. Hammond believed that gains from selection so founded would remain consistent when the individuals or their descendents were removed to more restrictive environments. His sustained interest in optimizing environment, both for performance itself and for the most effective selective breeding has been commended (Dickerson, 1962).
In animal production utilization of the most economically acceptable feeds and forages does not necessarily coincide with those which stimu- late the most desirable expression of traits.And Hammond failed to realize that interactions were possible between heredity and environment.
Wright (1939), among others of that time, recognized the non-additive
element of genetic and environmental effects on phenotype. It became evident that development of breeds and crosses for the most efficient production given the variety of feed resources available would require the examination of the suitability of particular breeds for the utiliza-
tion of specific feeds.
Thereafter, attempts were made to classify the alternative types of
genotype x environment interactions. The first attempt by Haldane (1946)
identified six interactions of four basic types which could result from two environments and two genotypes. They were: no interaction, a change in variability or magnitude of phenotypic difference, a change in rank or sign of the difference, and a change in variability and rank.
Assuming a limitation on which genotype would rank first and in which environment, Haldane showed that the number of possible types of inter- actions was the number of possible permutations of the remaining unique genotype and environment combinations. McBride (1958) graphed Haldane's six interactions. McBride (1958) and Dunlop (1962) classified four situations wherein interaction could occur. They were interactions of within population differences with small or intangible environmental differences, of within population differences with large differences in environment, of between population differences with small environmental differences and of between population differences with large environ- mental differences.
Small environmental differences are difficult to identify and have not been studied extensively. Interactions of within population diffe- rences with large environmental differences have beenstudied by com-
paring responses to selection in two environments. The most important
interaction in animal breeding occurs from between population differences
and large environmental differences. The economically important inter-
actions are those which involve a change in rank. An interaction which
changes the magnitude of diffetence between genotypes and not their
rankings results in a change in heritability and selection differentials
and thus alters the effectiveness of selection. But the consequences of
rank changes are more important. Studies have been conducted to determine the presence of genotype x environment interactions in sheep. However, the majority of the work done has been on wool traits and on Australian Merinos.
Dunlop (1963) found no strain x location or strain x year effects either in lambs born or lambs weaned in Australian Merino sheep. Pattie
(1965) found flock (wt. +, random bred, and wt. -) x year effects to be non-significant for fertility, lambs born and lambs weaned (as percents of ewes joined) and for prolificacy and survivability in Australian
Merinos. However, Dun et al. (1966) found significant strain x season interactions in Australian Merinos for fertility and prolificacy.
Carter et al. (1971), in an experiment with locations in Virginiaand
Quebec which had a one month lapse between their breeding seasons,ob- served significant breed cross x location interactions for poundsof
lamb weaned per ewe mated, lambing date and breed cross x year inter- actions for weight of lamb weaned per ewe which lambed. Lax and Turner
(1965) in another study on Australian Merinos found no significant
strain x location interactions for survival. Birth weight, which has an important effect on lamb survival (Malik andAcharya, 1972; Sidwell
and Miller, 1971; Eltawil et al., 1970) was not affected by sire xflock
interactions (Osman and Bradford, 1965), nor by strain x location or
strain x year interactions (Dunlop, 1963).
Dunlop (1963) found no significant strain x year or strain xloca-
tion effects on weaning weight in five strains of AustralianMerinos.
Sire x location effects were also non-significant onweaning weight from
studies on lowland and hill pasture by Osman and Bradford(1965). Carter et al. (1971), however, did find significant breed cross x location interactions for weight of lamb weaned per ewe mated and breed cross x year interaction for weight of lamb weaned per ewe whichlambed. Morley
(1956) found large and highly significant interactions between genotype and plane of nutrition for body weight at 12 months, using Australian
Merino half sibs. And Osman and Bradford (1967) found a significant interaction between heredity and ration energy for postweaning average daily gain but not for final weight using grade Targhee half sibs.
However, King and Young (1955) found these effects to be non-significant in a similar study using three different breeds, and again (King et al.,
1959) using twin lambs.
It is concluded here that large genetic differences, breeding sea- son differences, and measurable differences in plane ofnutrition have been the most important causes for genotype x environment interactions in the ewe production traits of interest in this study.
Location was documented by Carter et al. (1971) to have a signifi- cant effect on fertility, lambs born and lambs weaned per ewemated and on weight of lamb weaned per ewe which lambed.
Year was reported as a significant effect on fertility and lambs weaned per ewe mated by Sidwell, Everson and Terrill (1962), Sidwell and Miller (1971) and by Carter et al. (1971). Year was reported to have a significant effect on lambs born (Sidwell, Eversonand Terrill, 1962) and weight of lamb weaned per ewe mated and per ewe which lambed(Carter et al., 1971). The year effect on survival was reported to be signifi-
cant by Sidwell, Everson and Terrill (1962) but to benon-significant 8
by Sidwell and Miller (1971). Year was non-significant on survival to
14 days but significant from 15 days to weaning according to Malik and
Acharya (1972). The year effect on weaning weight is generally found to be significant (Sidwell, Everson and Terrill, 1964; Eltawil et al.,
1970; Vesely and Peters; 1972 and Holtmann and Bernard, 1969).
Age of dam has had an influence on prolificacy (Glimp, 1971;
Sidwell and Miller, 1971 and Sidwell, Everson and Terrill, 1962), on lambs weaned per ewe mated (Sidwell.and Miller, 1971; Sidwell, Everson and Terrill, 1962), on survival (Sidwell, Everson and Terrill, 1962;
Iwan, Jefferies and Turner, 1971), and on weaning weight (Sidwell,
Everson and Terrill, 1964; Ch'ang and Rae, 1970; Eltawil et al., 1970;
Holtmann and Bernard, 1969; Vesely and Peters, 1972; and Singh et al.,
1967).
Sex of the lamb has a significant effect on weaning weight
(Eltawil et al., 1970; Neville, Chapman and Pope, 1958; Sidwell, Everson and Terrill, 1964; Veseley and Peters, 1972).
Lamond et al.(1973) found lower fertility in ewes fed high and low energy diets than in pastured ewes due to low fertilization rates and low ovulation rates, respectively.The protein content of the ration did not influence fertility nor did it influence prolificacy in a study by
Torrell, Hume and Weir (1972). These authors attribute the increase in ovulation rate to high nutrient consumption of a composition compatible with weight gain. This flushing effect on ovulation rate was reported also by Bellows et al. (1963).
A significant effect of birth type on weaning weight has been shown by Sidwell, Everson and Terrill (1964); Neville, Chapman and
Pope (1958); Ch'ang and Rae (1970); Eltawil et al. (1970) and Veseley and Peters (1972). Ch'ang and Rae (1970) found that twins compared with singles exhibited post-weaning compensatory growth.
Lambing date reflects two things, the gestation length of the ewe, and breeding date which may indicate repeat breeding. Ewes bearing twins have a shorter gestation length (Dry, 1933; Thrift and Dutt, 1972) and ewes bearing males a longer gestation (Thrift and Dutt, 1972) but these differences were not significant. Thrift and Dutt (1972) found that when regressed within sex and birth type, the heavier lambs had highly significantly longer gestation lengths. No cause and effect relationship was postulated. Terrill and Hazel (1947) found that age of ewe was the most important non-hereditary source of variation in gestation length (older ewes having the longer gestations) and that early bred ewes tended to have longer gestations.
Breeding date has a significant effect on prolificacy but not on fertility (Glimp, 1971). Prolificacy increases to a peak in the middle of the breeding season and then declines (Glimp, 1971). Early birth dates have a positive and significant effect on weaning weight (Sidwell,
Everson and Terrill, 1964).
Sire breed was reported by Fahmy et al. (1972) to have had an im- portant effect on lambing rate, survivability and birth weight.
Important sire breed effects on weaning weight have been documented by
Eltawil et al. (1970), Karihaloo and Combs (1971), Neville, Chapman and
Pope (1958), Sidwell, Everson and Terrill (1964), Veseley and Peters 10
(1972) and Singh et al. (1967).
An important dam breed effect on prolificacy was demonstrated by
Glimp (1971), Sidwell, Everson and Terrill (1962) and Bellows et al.
(1963). Dam breed effects are significant and important on weaning weight (Eltawil et al., 1970; Karihaloo and Combs, 1971; Sidwell, Everson
and Terrill, 1964 and Veseley and Peters, 1972).
Heterosis has been expressed in higher fertility, prolificacy and
lambs weaned per ewe mated from crossbred than from purebred matings
(Sidwell and Miller, 1971). Heavier weaning weights due to heterosis were found by Sidwell, Everson and Terrill (1964), Iwan, Jefferies and
Turner (1971), Veseley and Peters (1972), Miller and Dailey (1951),
Singh et al.(1967) and Karihaloo and Combs (1971).
Reports of heterosis have not been consistent though. Differences
found between weaning weights by Bradley et al. (1972) and Holtmann and
Bernard (1969) were attributed to additive genetic variation and maternal
ability. No significant heterotic effect on survivability was found by
Malik and Acharya (1972).
Heterosis x environment interaction studies were not found in the
literature with perhaps one exception. A heterosis x age of dam inter-
action has been suggested by the work of Iwan, Jefferies and Turner
(1971), with crossbred versus purebred ewes on ewe production traits.
They found more heterosis in adult crossbred ewes for lambs born per ewe
joined, and more for two-year-old crossbred ewes for lamb survival,
lamb body weight and pounds of lamb weaned per ewe joined. 11
III. MATERIALS
Data for this study were 865 ewe production records recorded from fall 1969 through fall 1972, to include three lamb crops. The experi- mental design used was a three breed reciprocal cross repeated in two
environments. Each year, in each environment, approximately 144 ewes were mated. These 865 matings resulted in 757 lambings wherein 1263
lambs were born. One-hundred and eight matings did not result in pregnancy, and in three lambings none of the lambs were born alive.
Forty-seven ewes with living lambs failed to wean any of their lambs.
Source of Experimental Animals
The three breeds of sheep used were Hampshire, Suffolk and
Willamette. The Willamette strain had been developed over a fifteen
year period at Oregon State University, by crossing Border Cheviotand
Dorset Horn rams with Columbia ewes, making reciprocal crosses among
these crossbred progeny, and exercising selection within the resulting
closed population for 120 day weight, conformation and condition
(Bogart, 1961; Bogart, 1964). The Suffolks and Hampshires were also
from Oregon State University flocks. The selection criteria for the
Suffolks had been the same as for the Willamettes. The Hampshires had
been selected on pounds of lamb produced per ewe or pounds of lamb pro-
duced per unit of ewe metabolic weight.
Mating Scheme
Two rams of each breed were used in each environment each year. 12
Each ram was bred to eight ewes of each of the three breeds to produce the mating scheme shown in table I. Different rams were used for each breeding season. Each rams' semen was tested for adequate motility.
Ewes were rerandomized into breeding groups within environments each year. Culling of ewes was based primarily on poor health and infer- tility. Ewes which left the experiment were replaced from straightbred flocks maintained in conjunction with other experiments.
Environments
The two environments were on Oregon State University land in the
Willamette Valley, on the outskirts of Corvallis, Oregon.Corvallis is approximately 60 miles inland from the central part of Oregon's coast.
The Willamette Valley is between the Oregon Coast Range and the Cascade
Mountain Range at about 123 degrees 15 minutes west latitude, and 44 degrees 35 minutes north longitude. The elevation at Corvallis is 224 feet. The normal total precipitation is 40 inches per year. The dis- tribution is typified by a wet winter and a dry summer. Normal monthly rainfall totals range from seven inches in December to 0.3 inches in
July. Mean monthly temperatures range from 38°F in January to 66 °F in July, with record extremes of -14 °F and 107°F. The average date of the first fall frost is October 31, and the average date of the last
spring occurence is April 12.
The "Hill Pasture" environment was 215 acres of rolling, non-
irrigated pasture of subterranean clover, fescue and perennial ryograss,
typical of much of western Oregon foothill country. It was fertilized
on alternate years with single superphosphate (0-20-0) in thefall on 13
TABLE I. MATING SCHEME USED EACH YEAR
a South Farm
Breed of Sire Hampshire Suffolk Willamette
1 2 1 2 1 2
b d Hampshire 8 8 8 8 8 8 48 Breed of Suffolk 8 8 8 8 8 8 48 Dam Willamette 8 8 8 8 8 8 48
24c 24 24 24 24 24 144e a This mating scheme wasrepeated in the Hill Pasture environment b Number of ewes of eachbreed bred to each ram c Number of ewes bred toeach ram d Number of ewes of eachbreed in each environment e Number of ewes in eachenvironment 14
lower and open areas to encourage clover growth (250 pounds peracre) and was fenced into six unequal pastures. Pastures were grazed in an irregular rotational routine contingent upon pasture conditions.
The "South Farm" environment was 40 acres of level,irrigated pasture of orchardgrass, white clover, perennial ryegrass,subterranean clover, some alta fescue, and an invasion of meadowfoxtail, velvet grass and somebentgrass. It was fertilized every fall with single super- phosphate at about 250 pounds per acre, to promote clovergrowth, and during the spring of 1971 it was fertilized with nitrogen tostimulate
the grass growth. The sprinkler irrigation was started afterthe appear-
ance of the forage suggested the need for itand usually was done in
about three applications between June and August. It was fenced into
ten plots. Pastures were grazed in a modified rotational system,usually
following a short interlude after irrigation.
Management Practices
Ewes were bred from about September 10 to October21 in a forty day
breeding season each year, and lambing occurred fromlate January
through mid-March. The first year, 1969-1970, one-half of thebreeding
groups in each environment were flushedwith grain. Preliminary
analysis showed no influence on fertility orprolificacy from this treat-
ment, so flushing was excluded as a variablein subsequent years and
analyses. Ewes were kept on pasture in their respectivelocations all
year, except during lambing when the ewesfrom both environments were
in the university barn. Some years excess forage was harvestedoff part
of either location. Lambs were not creep fed. Lambs were slaughtered 15
when they reached 90 to 110 pounds or were weaned in early July and fed to 90 to 110 pounds.
All lambs were docked and male lambs were castrated at about two weeks of age with a knife. When lambs were about a month old they were vaccinated with enterotoxemia toxoid, strains C and D. Some years facial echthyma and selenium treatments were given. Thiabendazole was administered as a strategic anthelmintic at the start of the breeding season, sometimes before, and after the lambing season. Two shepherds were employed for the duration of the experiment. Management practices were held uniform over the three year period. Data were collected in a uniform format for the three years with weights taken as close to the same dates and/or intervals as possible. 16
IV. METHODS
As stated in the introduction, a primary purpose of this study was to determine the importance of genotype x environment interactions for ewe production traits. Genotype x environment interactions, if important, could lead to breed characterizations referenced to specific environ ments or to different expectations for the amounts of heterosis to be expressed. To quantify genotype x environment interactions, comparisons were made between scores and/or rankings of breeds and crosses in one location or year with their rankings in another for a given trait.
The average of the least squares means of the purebred progeny and the average of the least squares means of the crossbred progeny were com- pared to give measures of heterosis for each trait in each environment.
Comparisons of least squares means and measures of heterosis between, environments quantify the genotype x environment interactions found in this study. Significance of the genotype x environment interactions was determined by the F values of the two-, and the three-wayinteraction terms of sire and dam breeds with location and year in a least squares analysis of variance.
Other known factors which induce variation in the dependent variables studied were included in the statistical models or adjustments were made for them in the data before analysis. Inclusion of these known
effects in the model results in a more accurate partitioning of variance.
However, prior adjustment of the data can sometimes result in a more applicable analysis for some purposes. Such may be the case, when records
are commonly kept using industry correction terms. 17
In these analyses, year was included as a main effect since this results in the most accurate comparisons of breed effects and location
effects, the economically more important parts of this study. In
industry, comparisons are most often made within years so correction
factors would not be appropriate. The question of whether variation was within or between breeds necessitated the inclusion of sires as a main effect nested within breed of sire, location and year. Except in
the fertility analysis (which was done on percentages computed from breed-
ing groups, and not on observations from individual ewes) age of dam was
included as a main effect. Age interactions with location and year were
also included. These could result from differences between ages in
ability to cope with environmental stresses. Age interactions with
breed of ewe could result from different maturation rates and/or life
spans between breeds. The weight change of the ewe during the mating
season was included to determine and account for any flushing effect.
The number of days between the start of the mating season and a ewe's
lambing date reflects the number of estrous periods elapsed before con-
ception plus the rather constant gestation period. Repeat breeding is
a reproductive probleM which like fertility (which had already been
analyzed) is not highly heritable. So lambing date was included as a
continuous effect in the analysis of the ewes which lambed to determine
and account for its effects on ewe production. Variation attributable
to sex of lamb and age at weaning weight was accounted for by adjustment
of the data. Sex, at this time, is a random, uncontrollable effect.
Sex and lamb age at weaning are traditionally corrected for in industry. 18
Prior adjustment in this analysis, therefore, resultsin more economic relevance.
'Pounds of lamb weaned' was adjusted to a commonlamb age and sex by the following formula; Weaning Weight -BirthWeight) Adjusted Lamb Weight = (t X 135.6 + Birth Weaning Age Weight) X Sex Adjustment Factor (1) where 135.6 was the actual average weaning age in daysof all the lambs in this study. Sex adjustment factors used were 1.00 for wethers, and
1.03 for ewe lambs, industry correction terms from TheSheepman's
Production Handbook published by the Sheep Industry DevelopmentProgram
(Scott, 1970). Adjusted weights of all a ewe's lambs from each season were summed to give the 'pounds of lambweaned' figures.
Components of the Fertility Model
Fertility was considered to be the ability toconceive and carry at
least one lamb to full term. Included in the fertility model were
location, year, sire breed, and dam breed, plusall their two-, and
sire three- and four-way interactions. Also included was the individual
effect, within sire breed, location and year. The model used in the
analysis was:
= a + Li + Y. + S + + Rm/Sk/Li/Yj (LS) + (YS)jk (2) Fijklmn k + (LYS)ijk + (LD)ii + (YD)ji +(LYD)iji + (SD)kl (LSD)iki + (YSD)jkl
(LYSD) ijkl eijklmn
F.. represents the observed fertility(measured as the percent of ijklmn .th those ewes mated which actually lambed, of thebreedirg group in the 19
th th location (L) in the (Y), with the k sire breed (S), with the th th 1 dam breed (D) and the m individual ram unique CR) to a location and year, where 'a' is the population mean value common to all such obser- vations. The e term accounts for random differences between ijklmn observations and includes variation unaccounted for by the effects in- cluded in the model.
Components of the Model Analyzed on a Per Ewe Mated Basis
A least squares analysis for unequal subclass numbers was used to analyze the data for this study. Dependent variables analyzed on the basis of all ewes mated were: the number of lambs born alive, the total number born, the number weaned, and the pounds of lamb weaned per ewe.
Thus zero observations were included for all variables. Independent variables included in the model were: sire breed, dam breed, location, year, the individual sire effect nested within sire breed, location and year, the age of the dam, the weight change of the dam during the mating season as a continuous variable, and the interaction terms judged to be appropriate as indicated in the introduction of this chapter. The model used in the analysis was:
Xijklmnop = a +Si + Dj + (SD)ij + Lk + (SL)ik + (DL)jk + (SDL)ijk + Y1(3)
(SDY)iji + Rm/Si/Lk/Yi + An + (SY)il (DY)j1 (LY)kl +
(YA) + (DA). + W + e. ln 3n o ijklmnop Xijklmnop represents the observed value for the dependent variable (born th alive, born total, number weaned, pounds weaned) for the ewe with the o weight change (W) of the jth dam breed (D) in thenthage class (A), bred 20
th to the m (R) of theithsire breed (S), in thelthyear (L), in the th k location (L), where 'a' is the population mean value common to all such observations. The term represents a random error peculiar eijklmnop to each observation and includes variation not attributable to theeffects included in the model.
Components of the Model Analyzed on a Per Ewe Lambing Basis
The least squares analysis for unequal subclass numbers was used to analyze the data for this study also. Dependent variables analyzed on the data from only thoseewes which lambed were: the number of lambs born alive, the number born total, the percent born alive, the number weaned, the percent weaned of those born alive and the pounds weaned per ewe lambing.
Independent variables included in the model were: sire breed, dam breed, location, year, the individual sire effect, age of the dam, weight change of the ewe during the mating season as a continuous vari- able, the number of days between the start of the mating seasonand the ewe's lambing date also as a continuous variable, and those interactions judged to be appropriate.
At this point the age interactions had not been shown to haveimpor-
tance in the analysis of traits for all ewes mated. It was decided that
the differences betwen locations and years in so temperate aclimate
would not likely provoke differences in age abilities to cope with
stress, especially over the small age interval in thesedata. Thus age
interactions were eliminated, since inclusion of non-significantfactors
in a model may introduce biases in the important effects and takes away 21
significance from the other effects (Harvey, 1960). The model used in this analysis was;
Si + D. * (SD)ij + Lk (DL)jk + (SDL)ijk + (4) Xijklmnopq a (SL)ik yl + (sY)ii (DY)ji + (LY)iii + RIII/Si/Lk/Y1 + An + Wo Gp eijklmnopq
X. represents the observed score for the dependent variable for ijklmnopq th th the ewe with the o weight change (W) and the p days to lambing (G),
.th th of the 3 dam breed (D), in the n age class (A), bred to the mthm
. th th th ram (R) of the sire breed (S), in the 1 year (Y), in the k location (L), where 'a' is the population mean common to all such obser- vations and e accounts for random differences between observa- ijklmnopq tions and also includes variation not attributable to the effects in- cluded in the model. 22
V. RESULTS AND DISCUSSION
The small amount of variation in livability (percent born alive of lambs born total) suggested the omission of it and of lambs born alive in subsequent discussions, and lambs born total per ewe which lambed was taken as the best reflection of prolificacy. The mean squares for the eight remaining analyses of variance described in the methods section and their tests of significance are presented in tables II, III and IV.
The first is the analysis of fertility, then the three analyses based on data from all the ewes mated (M), followed by the four analyses based on only those ewes which lambed (L). Least squares means are given in the appendices, A through D. Means for each dependent variable are given for the effects of sire breed and each dam breed and for each of the specific crosses. They were computed by adding the deviations for the appropriate sire breed, dam breed and interaction combination to the overall least squares mean which had been adjusted, using the appro- priate partial regression coefficients, to the mean values of the con tinuous variables. The first dependent variable shown is fertility
(percent which lambed of ewes mated) in appendix A. Then means for lambs born, lambs weaned, and pounds of lamb weaned based on all ewes mated(M)
are given in appendix B. The three dependent variables just mentioned
plus percent weaned (survivability) are then given based only on those
ewes which lambed (L) also in appendix B. The means for all the depen-
dent variables listed except fertility are repeated, presentedwithin
years and within environments in appendix C. Location and overall means
for each trait are shown in appendix D. Heterosis, calculated from these 23
TABLE II. LEAST SQUARES ANALYSIS OF VARIANCE FOR EWE FERTILITY (%)
Source of Variation d.f. Mean Squaresa
Locations (L) 1 48.0
Years (Y) 2 212.7
LY 2 6.3
Sire Breed (S) 2 49.8
LS 2 29.7
YS 4 228.5
LYS 4 59.5
Dam Breed (D) 2 205.4
LD 2 301.4
YD 4 186.9
LYD 4 113.9
SD 4 148.6
LSD 4 175.6
YSD 8 119.0
LYSD 8 72.3
Nested Ram Effect 18 164.9
Error 36 97.9
a None of the effects was significant 24
TABLE III. LEAST SQUARES ANALYSESOF VARIANCE FOR EWE PRODUCTIVITY BASED ON EWES MATED(M)
Mean Squares Source of Lainbs Lambs Pounds Variation d.f. Born Weaned Weaned
Sire Breed (S) 2 .14 .17 1755.
Dam Breed CD) 2 .73 .67 14531.*
SD 4 .68 .58 10088.*
Location (L) 1 1.05 .39 24361.**
SL 2 .97 1.14a 7041. a DL 2 .58 1.19 10122.* a SDL 4 .86 1.10 7959.*
Year (Y) 2 3.72** 2.57** 13684.*
SY 4 1.51** .91 6661.a
DY 4 1.02a .82 6312. a LY 2 .09 1.12 24820.**
SDY 8 ,65 .80a 4318.
Nested Ram Effect 18 .62 .47 3279.
Dam Age 4 3.63** 3.84** 30173.**
Dam Weight Change 1 .02 .91 20999.*
Error 804 .52 .49 3321.
a P<.10 **P<.05 *P<.01 25
TABLE IV. LEAST SQUARESANALYSES OF VARIANCE FOR EWE PRODUCTIVITY BASED ONEWES WHICH LAMBED CL)
Mean Squares Source of Lambs LambS PerCent Pounds Variation d.f. Born Weaned Weaned Weaned
Sire Breed 2 .05 .03 272 911
Dam Breed 2 .18 .14 243 4064
SD 4 ,10 .30 1176 4437
Location 1 6,45** ,08 10162** 13157*
SL 2 .18 .35 339 871
DL .02 .28 399 2173
a a SDL 4 .55 .7979a 1062 4661
Year 2 4.25** 5.43** 481 40426**
SY 4 .85* .48 356 2387
DY 4 .37 .33 1459 2533
LY 2 .08 1.31* 2970* 26421**
SDY 8 .28 .55 888 3055
Nested Ram 18 .18 .26 834 1450 Effect
Dam Age 4 3.15** 3.53** 934 25114**
DamWeight a 1 1.54* .04 2782 1739 Change Lambing Date 1 1.80* 2.26* 267 2902
Error 695 .29 .35 811 2186
a P<.10 *P<.05 **P<.01 means by dividing the crossbred mean by the purebred mean, subtracting one and multiplying by 100, are given in appendix E.
Environmental Effects and Interactions
The least squares deviations for all the environmental effects tested are tabulated in table V.
Location was significant for pounds weaned (M), lambs born (L),
(or prolificacy) percent weaned (L)(or survivability), and pounds weaned
(L). Except for prolificacy the performance on hill pastures was better than on irrigated pastures. Significant location effects (Virginia` vs.
Quebec, Canada) were found by Carter et al. (1971) for fertility, lambs born (M), lambs weaned (M), and weight of lamb weaned (L). Ewes on valley pastures bore more lambs but weaned fewer than did ewes on the hill pastures based on all the ewes mated. This magnitude of difference in survivability indicates that the two locations also had different levels of stress. In fact, the valley lambs were kept in confinement in the lambing barn longer than hill pasture lambs so had greater exposure to disease and parasitism, notably pneumonia and coccidiosis.
Also single lambs have better survivability from birth to weaning than have twin lambs. Birth type has a highly significant effect on birthweight (Karihaloo and Combs, 1971) and birth weight was the only
significant factor on survival to 14 days found by Malik and Acharya
(1972). Sidwell and Miller (1971) reported that single lambs survived better from birth to weaning than twin lambs. The lower survivability,
in the valley pasture may therefore have been partially due to the TABLE V. ENVIRONMENTAL EFFECTS
Per Ewe Mated (M) Per Ewe Which Lambed(L) Lambs Lambs Pounds Lambs Percent Pounds Effect Fertility Born Weaned Weaned Born Weaned Weaned Weaned Valley -.7 .05 -.03 -8,1 .12 .01 -4.8 -5.5
Hill .7 -.05 .Q3 8.1 -.12 -.01 4,8 5.5
1969-1970 2.8 .33 .33 32.9 .14 .16 1.4 10.6
1970-1971 -1.9 -.30 -.27 -22.0 -.13 -.14 -1.5 -9.8
1971-1972 -.9 -.03 -.06 -10.9 -.01 -.02 .1 -.8
2 yr. old -.25 -.29 -27.8 -.23 -.27 -4.4 -23.3
3 yr. old -.08 -.07 -4,0 -.10 -.08 1.1 -2.6
4 yr. old .11 .07 3.4 .13 .10 ,2 7.9
5 yr. old .13 .12 11,2 .05 ,07 -.2 7.4
6 & 7 yr. old .08 .17 17.2 ,14 .18 3.4 10.6 a Wt. Change .00 .00 .6 -.01 -,00 ,2 .2
b Lambing Date ,01 .01 .1 .2
Overall Mean 88.6 1.46 1.25 113.4 1.64 1.40 87.2 123.8 a The average weight change for ewes mated was 10.6 pounds. The average weight change for only those ewes which lambed was 10.9 pounds. bTheaverage lambing date was 159.6 days after the start of the mating season. 28
high prolificacy. Only on the basis of ewes which lambed did valley ewes wean more lambs and then they still weaned fewer pounds of lamb
(ML) than did hill pasture ewes. The larger proportion of twin lambs on the valley pasture with their consequentially depressed rateof gain,
(Sidwell, Everson and Terrill, 1964, Neville, Chapman and Pope, 1958,
Chiang and Rae, 1970, Eltawil et al., 1970, Veseley and Peters, 1972) and lower survivability culminated in the observably poorer pounds weaned figures on both per ewe mated and per ewe which lambed bases.
Year was significant for all traits except fertility and surviv- ability. Apparently year differences were not great enough to provoke differences in survivability. Sidwell and Miller (1971) and Carter et al.
(1971) found significant year effects on fertility and lambs weaned (M).
Carter et al. (1971) also reports significant year effects on lambs born
(M) and weight of lambs weaned (OL). Year effects on survivability tested by Sidwell and Miller (1971) were not significant. Malik and
Acharya (1972) did find significant year effects on survival after 15 days but not on survival to 14 days. Important year effects on weaning weight have been reported by Sidwell, Everson and Terrill (1964),
Eltawil et al. (1970), Veseley and Peters (1972), and Holtmannand
Bernard (1969).
Dam age was highly significant for all traits except survivability and fertility was not tested. Over the age interval studied, most of the production traits tested increased steadily. A decrease in lambs born per ewe mated while lambs born per ewe lambing continued toincrease,
indicated a decrease in fertility in the oldest age group (six and seven 29
year old ewes). The significance of this difference was not tested be- cause fertility was estimated as the percent lambing in each breeding group and did not include individual information on the ewes. Since barrenness was one of the bases for culling during the experiment, the age effect on fertility was probably even more pronounced than these re sults suggest. Sidwell and Miller (1971) and Glimp (1971) reported significant ewe age effects on prolificacy. Sidwell and Miller (1971) also found significant ewe age effects on lambs weaned (M). An important ewe age effect on lamb survivability was found by Iwan,Jefferies and
Turner (1971). Ewe age has an important effect on lamb's weaning weight
(Eltawil et al., 1970, Sidwell, Everson and Terrill, 1964, Veseley and
Peters, 1972, Ch'ang and Rae, 1970, Singh et al., 1967, Holtmann and
Bernard, 1969).
The weight change of the dam during the mating season was signifi cant for pounds weaned (M), and lambs born (L). It approached signifi-
cance for percent weaned (L). The effect of weight change on prolificacy was negative, opposite to what is usually reported inthe literature,
but the effect was small and not important. Torrell et al. (1972) found
a beneficial effect of weight gain during the mating season onprolificacy.
The weight change, as measured in this study, most likely was not agood
indication of flushing effects.Perhaps it was measured over too long
a period or should have been calculated in aperiod before or in the
early part.of the mating season. The effect of weight change on pounds
weaned and percent weaned was positive. A ewe which gave a lot of milk
could be expected to gain during the subsequent mating season in re-
cuperation from the strain of lactation the former season, and eweswhich 30
milk well are expected to wean more of their lambs and heavier lambs.
The lambing date was significant for lambs born (L), and for lambs weaned (L). The effect was positive on both variables. Prolificacy in creases to a peak in the middle of the breeding season (Glimp, 1971,
Hammond, 1944). A positive effect of early lambing dates on weaning weights has been reported as a significant effect in the literature but was not significant in this study (Sidwell, Everson and Terrill, 1964).
The environmental interaction of location x year was significant
for lambs weaned (L), percent weaned (L), and pounds weaned (MIL).
Carter et al. (1971) found highly significant location x year interactions
for lambs born (ML), lambs weaned (OL), and weight of lamb weaned
(ML). The valley pasture consistently showed more variation for these
traits between these three years than did the hill pasture. The valley main effect was negative for these traits except for lambs weaned (L).
The sheep in the poorer valley pasture may have been less buffered to
environmental fluctuations. For example, a wet May which depresses wean-
ing weight due to exposure and parasite problems (Singh and Nelson, 1955)
would clearly insult the lambs on the more heavily stocked and unsheltered
valley pasture more than the hill lambs. The significant effects of
location, year and their interaction are plotted in figures 1 through 4.
As well as a change in the amount of year variation between locations,
year rank changes occurred in survival and in pounds of lamb weaned (L)
between locations. A change of location rankings between years also
occurred for all these traits except survival which was consistently
better in the hill pasture. These differences in variability and/or 31
.01 Year Mean
Valley Hill Location
Figure 1. Location x Year Interaction for Lambs Weaned per Ewe Lambing Valley' Hill Location
Figure 2. Location x Year Interaction for Percent Weaned Valley Hill Location
Figure 3. Location x Year Interaction for Pounds Weaned perEwe Mated Valley Hill Location
Figure 4. Location x Year Interaction for Pounds Weaned per Ewe Lambing 55
rank of locations in years accounts for the interactions.
Breed Effects and Breed x Environment Interactions
Sire breed effects and dam breed effects are tabulated in tableVI.
Dam breed was a significant effect on pounds weaned (M), the singlebest measure of ewe production, Hampshire ewes mated weaned 102,7 pounds,
Suffolks 118.3 and Willamettes 119.2 pounds. Hampshire ewes had good fertility, were the most prolific and had the poorest survivability, possibly due in part to their high prolificacy. They weaned the most lambs (L), but the least pounds of lamb (MEL). Suffolk ewes had the poorest fertility, the poorest prolificacy, and possibly as a conse- quence the highest survivability. Suffolk ewes weaned the most pounds per ewe which lambed. Willamette ewes had the highest fertility, slightly poor prolificacy and survivability but weaned an average pounds of lamb per ewe which lambed. Important dam breed effects on weaning weights of lambs were reported by Eltawil et al 1970,
Karihaloo and Combs, 1971, Sidwell, Everson and Terrill, 1964and Veseley and Peters, 1972. Pounds weaned (M) in the comparison of thesethree breeds has a very large dam breed effect as illustrated byfigure 5. This
is probably responsible for the significant dam breedeffect while sire breed was not significant.
Neither the sire breed nor the individual ram effectnested within
location within year was significant for any of thetraits examined.
Hampshire rams had the poorest effect on fertility, average on pro-
lificacy and survivability and an average to poor effect onpounds TABLE VI. BREED EFFECTS
Dam effect Sire effect a
Fertility 1.2 -2.8 1.6 -1.3 .2 1.1
Mated
Lambs Born -.07 -.00 .07 .03 -.01 -.02
Lambs Weaned -.07 .02 .05 .03 -.01 -.02
Pounds Weaned -10.7 4.9 5.8 1.2 2.0 -3.2
Lambed
Lambs Born .03 -.02 -.01 -.00 -.01 .02
Lambs Weaned .03 -.01 -.02 .01 -.01 .01
Percent Weaned -.9 1.1 -.2 .2 1.1 -1.3
Pounds Weaned -4.1 4.2 -.0 -1.0 2,4 -1.4
a Hampshire b Suffolk c Willamette 9.0, 37
7.0.
.40, Dam Effect 44044wIlw .4044°' 5.0`
3.0-1
1.0" 0 Mean pi
3.0 Sire Effect
- 5.0.
7.o-
-9.0
-11.0
Hampshire Suffolk Willamette Breed Figure 5. Breed Effects on Pounds Weaned per Ewe Mated 38
weaned. Suffolk rams had an average effect on fertility, poor on pro- lificacy, and the best effects on survivability and pounds weaned (M&L) the two traits where sire effects were greatest. Willamette rams had the best effect on fertility and prolificacy, and the poorest on surviv7 ability and pounds weaned (M$L), Significant/sire breed effects have been found on lambing rate and survivability, possibly through the sig.:. nificant effect on birth weight (Fahmy et al 1972). Significant sire breed effects on weaning weight have been found by Eltawil et al.,
1970, Karihaloo and Combs, 1971, Neville, Chapman and Pope, 1958, Sidwell and Everson and Terrill, 1964, Veseley and Peters, 1972 and Singh et al.,
1967.
Dam breed x location for pounds weaned (M), and sire breed x year for lambs born (NM) were significant breed x environment interactions.
Carter et al.(1971) reported a significant location x breed cross interaction for pounds weaned (M) and a significant year x breed cross interaction for pounds weaned (L). Sire and dam breed effects were not differentiated by Carter et al.(1971). In figures 6, 7 and 8 these effects are graphed. The differences between dam breed effects increased in magnitude between locations and even showed a change in rank. There was more opportunity for genetic difference to be expressed and a larger magnitude of difference between dam breeds was produced in the better location. The greatest part of this interaction is due to a conspicuous change in the performance of the Willamette ewes and their particular suitability to the hill pasture. The overall ranking of ewe breeds for pounds weaned (M) holds true in the better hill pasture. But Suffolk 39 Willamette
10.0 Suffolk
Valley Hill Location
Figure 6. Dam x Location Interaction for Pounds Weaned per Ewe Mated 40
-o 0
0 ua .10 0
Willamette 0;-4 ca Suffolk Mean
4-i 0 0
0
Figure 7 Sire x Year Interaction for Lambs Born per Ewe Mated fiec, e* 4, ,?.?.
1
'?c, <3 42
ewes performed better in the valley than Willamette ewes. In the valley,
Hampshire ewes mated weaned 98.8 pounds, Suffolks 113.0 pounds, and
Willamettes 104.2 pounds, On the hill, Hampshire ewes mated weaned
106.7 pounds, Suffolks 123.5 pounds and Willamettes 134.2 pounds.
It is of particular interest to note that the Willamette breed was developed in hill pastures for use in Western Oregon (Bogart, 1961;
Bogart, 1964). The Willamette breed is not an economically important one due to relatively small numbers, but its inclusion in this experi- ment is of unique importance. What it has demonstrated suggests that perhaps a unique adaptation to the hill environment has been produced in the Willamette breed through selection.
Rankings and magnitudes of differences between sire breeds for number of lambs born (M&L) changed between years. Interaction effects alone however, were in no case larger than one tenth of a lamb. Thus the importance of this interaction is questionable.
Heterosis and Heterosis x Environment Interaction
The heterosis of these two breed cross lambs and its interaction with location both had significant effects on pounds of lamb weaned
(M). They are illustrated in figure 9 and 10. Sometimes more extensive, although not always consistent, reports of heterosis are found in the literature. Sidwell and Miller (1971) generally found higher fertility, prolificacy and lambs weaned (M) for crossbred than for purebred matings.
With the exception of the Willamette x Hampshire cross lambs when in the hill pasture, all crosses were equal or superior to the midparent average of their component breeds. With the added exceptions of 43
14.7% Heterosis Overall
Alamammo Midparent average H Hampshire S Suffolk W Willamette
11.8% 16.7%
H HWW HHSSWWSS Sire Breed H W H W H S H S WS W S Dam Breed
Figure 9. Means of Specific Crosses for PoundsWeaned per Ewe Mated 44
Overall 14.7% Valley 31.3% Valley Hill 8.2% Hill H Hampshire S Suffolk H X W H X S S X W W Willamette Valley 31.4% 34.8% 28.7% Hill -1.9% 1.5% 7.8%
ca 0)
II
Sire H H WW IIHSS Dam W H S H S
Figure 10. Heterosis x Location Interaction 45
Willamette x Hampshire in the valley pasture, Hampshire x Suffolk,
and Suffolk x Hampshire in the hill pasture, all crossbred matings were superior to the better purebred mating of their componentbreeds
for this trait, These exceptions are due to the negative effect of
the maternal ability of the Hampshire breed (tableV1), and to the
relative lack of genetic diversity between the two blackface breeds,
Suffolk and Hampshire. A smaller gene frequency difference between
breeds results in less heterosis. Willham (1969) shows that the
squared difference between the gene frequency of a given allele in two
breeds times the degree of dominance is the amount of heterosis expressed.
Gene frequency differences between breeds are a function of their
diversity. Two breeds of as similar origin as Hampshire and Suffolk
could be expected to show relatively little heterosis.
Heterosis did not consistently produce a crossbred mating advantage
in all cases in the location which had a positive main effect for this
trait as is illustrated in figure 10. Figure 10 shows breed group means
from valley pastures with blacked bars and those from hill pastureswith
open bars. Notice that the superiority of the crossbred matings was
far more apparent in the poorer valley location. Purebred matings
suffered more from the depressing effect of the poorer location thandid
crossbred matings. Also notice that the crossbred matings in the poorer
location much more nearly approximated the crossbred matings in the
better location than did purebred matings approximate the purebred mat-
ings in the better location. Crossbred lambs were better able to over-
come their environmental disadvantage. A general crossbred advantage
for weaning weight of lambs from purebred ewes has been reported by 46
Sidwell, Everson and Terrill (1964), Iwan, Jefferies and Turner (1971),
Veseley and Peters (1972), Miller and Dailey. (1951), Singh et al. (1967) and Karihaloo and Combs (1971), However, Neville, Chapman and Pope
(1958), Bradley et al, (1972) and Holtmann and Bernard (1969) attributed differences in weaning weights between breeding groups solely to additive genetic and/or maternal effects for crossbred lambs from purebred ewes and found no important heterOtic effects. Malik and Acharya (1972) found no significant heterotic effects on survival. Possibly differences in the genetic diversity of breeds or the quality of the environments contributed to these discrepancies. 47
VI. SUMMARY AND CONCLUSIONS
Eight-hundred and sixty -five ewe production records taken from three lamb crops, spring 1970, 1971 and 1972 were analyzed to study breed effects, heterosis effects, environmental effects and interactions between them, and breed and heterosisx environment interaction effects on ewe production traits.
Each of the three years 288 Hampshire, Suffolk and Willamette ewes from university flocks were mated in a three breed reciprocal cross, half in each of two locations. The resulting 757 parturitions produced
1263 lambs. Two rams of each breed were used at each location each year, and ewes were rerandomized each year. The two locations were near
Corvallis, Oregon, 60 miles inland at 224 feet elevation, in a temperate climate with a wet winter and a dry summer and with normal total precipitation 40 inches per year. One location was level irrigated valley bottom land, the other non-irrigated rolling hill pasture.
Least squares analyses of variance were computed on ewe production traits which included fertility, lambs born, percent survival to weaning, lambs weaned and pounds of lamb weaned. Main effects in the models were location, year, sire breed, dam breed, sires nested within locations within years, dam age, dam weight change during the mating season and lambing date. The two- and three-way interactions between sire breed, dam breed, location and year were also tested. Weaning weights of lambs were corrected for age by adding the lamb's own birth weight to its daily gain to weaning multiplied by the experimental average weaning age. Weaning weights were corrected for sex by-multiplicative industry 48 adjustment factors.
None of the effects tested in the fertility model showed significance.
On the other traits, year was significant except for survival.
Apparently year differences were not severe enough to provoke differences in survivability. Location was significant for pounds weaned per ewe mated and for lambs born, percent weaned and pounds weaned per ewe lambing. The main effects of the hill land were superior for these traits with the exception of prolificacy. Ewes on valley pastures had more lambs but weaned fewer than did ewes on hill pastures based on all the ewes mated. This magnitude of difference in survivability indicates different levels of stress. Valley lambs were kept in confinement in the
lambingbarn longer so had greater exposure to disease and parasitism, notably pneumonia and coccidiosis. The depressing effect of multiple birth types and of parasite and disease problems on survival, and on growth to weaning in the more heavily stocked and unsheltered valley pasture resulted in the lower pounds of lamb weaned shown. The
interaction between location and year was significant for lambs weaned,
percent weaned and pounds of lamb weaned per ewe which lambed and for pounds weaned per ewe mated. The interaction resulted from greater variation in the valley between these three years.
Dam age was significant except for fertility and survival.
Production generally increased with age. A decrease in fertility in the
oldest age group was indicated but not tested. Fertility was estimated
as the percent lambing in each breeding group and did not include individ-
ual information on the ewes. A decrease in lambs born per ewe mated
while lambs born per ewe lambing continued to increase indicated a 49 decrease in fertility in the oldest age group (six and seven year old
ewes). Since barrenness was one of the bases for culling during the
experiment, the age effect on fertility was probably even more pro- nounced than these results suggest. Dam weight change was significant
for pounds weaned per ewe mated and for prolificacy. Pounds weaned
increased with weight gain while prolificacy decreased. Lambing date was significant for prolificacy. Prolificacy increased from the start
of the mating season.
Dam breed and dam breed x location interaction were significant for
pounds of lamb weaned per ewe mated. This interaction is due almost
entirely to a disproportionately large increase in the Willamette ewe
production in the hill pasture. Ewes ranked Willamette (119.2 pounds),
Suffolk (118.3 pounds) and Hampshire (102.7 pounds) overall. The
ranking held true on the more productive hill location but with scores
of 134.2 pounds, 123.5 pounds and 106.7 pounds respectively. On the
valley the Suffolks surpassed Willamettes (113.0 to 104.2 pounds) and
Hampshires remained the poorest producers (98.8 pounds). Since the
Willamette breed was developed on hill land the interaction is of
particular significance. It suggests that perhaps a unique adaptation
to the hill land has been produced in the Willamette breed through
selection. Sire x year was significant for lambs born but the effect
was small and unimportant. The interaction effects alone were in no
case larger than one tenth of a lamb.
Heterosis and heterosis x location interaction were significant
for pounds weaned per ewe mated. The location with the poorer main 50
effect exhibited the greatest heterosis. Crossbreeding advantage was
least in the more closely related Hampshire and Suffolkcrosses and
the crosses of Hampshire ewes which had the poorest maternal effect with
Willamette rams which had the poorest sire breed effect.
Genotype x environment interactions were important effects in these analyses, causing the significant change of ranking of dam breed and breed crosses between the two environments for pounds weanedper ewe mated and the significant change of ranking of sire breeds for lambs born. Heterosis was 31.3% for pounds weaned per ewe mated on the valley pasture and 8.2% on the hill pasture. These results suggest that heterosis is increased by suboptimal conditions for the expression of a given trait. 51
Bibliography
Bellows, R. A., A. L. Pope, A, B. Chapman and L. E. Casida. 1963. Effect of level and sequence of feeding and breed on ovulation rate, embryo survival and fetal growth in the mature ewe. J. Anim. Sci. 22:101.
Bogart, Ralph. 1961. Development of a sheep for western Oregon. Sheep and Wool Day, Misc. Paper 121, Ore. Agr. Exp. Sta., O. S. U., Corvallis.
Bogart, Ralph. 1964. Lamb production from the Willamette breed of sheep. Sheep and Wool Day, $pec. Rpt. 182, Ore. Agr. Exp. Sta., O. S. U., Corvallis.
Bradley, B. P., A. B. Chapman, A. L. Pope and C. 0. Rydberg. 1972. Two- and three-way crosses estimating combining ability of Suffolk, Targhee and Shropshire breeds of sheep. J. Anim. Sci. 34:541.
Carter, R. C., G. M. Carman, F. S. McClaugherty and P. S. Haydon. 1971. Genotype-environment interaction in sheep. I. Ewe productivity. J. Anim. Sci. 33:556.
Ch'ang, T. S. and A. L. Rae. 1970. The genetic basis of growth, reproduction and maternal environment in Romney ewes. I. Genetic variation in hogget characters and fertility of the ewe. Aust. J. Agric. Res. 21:115.
Clarke, H. G. 1963. Commercial Sheep Management. Crosby Lockwood & Son, London.
Dickerson, G. E. 1962. Implications of genetic-environmental interaction in animal breeding. Anim. Prod. 4:47.
Dry, F. W. 1933. A note on the length of estrous cycle and duration of pregnancy in Romney ewes. New Zealand J. Agr. 47:386.
Dun, R. B., R. Alexander and M. D. Smith. 1966. An interaction between strain of Merino ewe and season of joining. Proc. Aust. Soc.Anim. Prod. 6:66. (A.B.A., 35:No. 456).
Dunlop, A. A. 1962. Interactions between heredity and environment in the Australian Merino. I. Strain x location interaction in wool traits. Aust. J. Agric. Res., 13:503.
Dunlop, A. A. 1963. Interactions between heredity and environment in the Australian Merino. II. Strain x location interactions in body traits and reproductive perforMance. Aust. J. Agric. Res. 14:690. 52
Eltawil, E. A., L. N. Hazel, G. M. Sidwell and C. E. Terrill. 1970. Evaluation of environmental factors affecting birth, weaning and yearling traits in Navajo sheep. J. Anim. Sci. 31:823.
Fahmy, M. H., C. S. Bernard, J. P. Lemay and M. Nadeau. 1972. Influence of breed of sire on the production of light and heavy market lambs. Can. J. of Anim. Sci. 52:259.
Glimp, H. A. 1971. Effect of breed and mating season on reproductive performance of sheep. J. Anim. Sci. 32:1176.
Haldane, J. B. S. 1946. The interaction of nature and nuture. Ann. Eugen. (Lond.) 13:197. (A.B.A. 15:66).
Hammond, J., Jr. 1944. On the breeding season in the sheep. J. Agr. Sci. 34:97.
Hammond, J., J . Edwards and A. Walton. 1941. Animal breeding in relation to environmental conditions. J. Royal Agri. Soc. of England. Vol. 142.
Hammond, J. 1936. Environment conditions and livestock breeding. Problems of Animal Husbandry (Moscow) No. 8.
Harvey, W. R. 1960. Least-squares analysis of data with unequal subclass numbers. A.R.S. 20-28. U.S.D.A.
Holtmann, W.B. and C. Bernard. 1969. Effect of general combining abilityand maternal ability of. Oxford, Suffolk and North Country Cheviotbreeds of sheep on growth performance of lambs. J. Anim.Sci. 28:155.
Iwan, L. G., B. C. Jefferies and Helen Newton Turner. 1971. Estimation of heterosis in Merino x Corriedale crosses with sheep. Aust. J. Agric. Res. 22:521.
Karihaloo, A. K. and W. Combs. 1971. Some maternal effects on growth in lambs produced by reciprocal crossbreeding between Lincoln and Southdown sheep. J. Anim. Sci. 51:511.
King, J. W. B. and G. B. Young. 1955. A study of three breeds of sheep wintered in four environments. J. Agr. Sci. 45:331.
King, J. W. B., J. H. Watson and G. B. Young. 1959. Genotype-environ- ment interactions in the wintering of lambs. J. Agric.Sci. 53:156.
Lamond, D. R., J. R. Hill, Jr., W. C. Godley, S. W. Kennedy and R. G. Gaddy. 1973. Influence of nutrition on ovulation and fertiliza- tion in the Rambouillet ewe. J. Anim. Sci. 36:363. 53
Lax, J. and Helen Newton Turner. 1965. The influence of various factors on survival rate to weaning of Merino lambs. I. Sex, strain, location and age on ewe for singleborn lambs. Aust. J. Agric. Res. 16:981. (A.B.A., 34:No, 1294)
McBride, G. 1958. The environment and animal breeding problems. Anim. Breed. Abstr. 26:349.
Malik, R. C. and R. M. Acharya. 1972. A note on factors affecting lamb survival in Indian sheep. Anim. Prod. 14:123,
Miller, K. P. and D. L. Dailey. 1951. A study of crossbreeding sheep. J. Anim. Sci. 10:462.
Morley, F. H. W. 1956. Selection for economic characters in Australian Merino sheep. VII. Interactions between genotype and plane of nutrition. Austr, J. Agr. Res. 7:140.
Neville, W. E., Jr,, A, B. Chapman and A. L. Pope. 1958. Comparison of lambs from Western (Columbia-Rambouillet) ewes and sired by rams of four down breeds. J. Anim. Sci. 17:763.
Osman, A. H. and G. E. Bradford. 1965. Effects of environment on phenotypic and genetic variation in sheep. J. Anim. Sci. 24:766.
Osman, A. H. and G. E. Bradford. 1967. Genotype-environment interaction and compensatory growth in sheep. J. Anim. Sci. 26:1239.
Pattie, W. A. 1965. Selection for weaning weight in Merino sheep. 2. Correlated responses in other production characters. Aust. J. Exp. Agr. Anim. Husb. 5:361. (A.B.A., 34 No. 1295).
Scott, George, E.(ed.) 1970. The Sheepmanls Production Handbook. Sheep Industry Development Program, USDAARS.
Singh, 0. N. and R. H. Nelson. 1955. Effect of certain environmental factors on reproductive performance of ewes and growth of lambs. Bull., Mich. Agr. Exp. Sta., E. Lansing 38:192.
Singh, B. P., W. E. Rempel, D. Reimer, H. E. Hanke, K. P. Miller and A. B. Salmela. 1967. Evaluation of breeds of sheep on the basis of crossbred lamb performance. J. Anim, Sci. 26:261.
Sidwell, G. M., D. O. Everson and C, E. Terrill. 1964. Lamb weights in some purebreeds and crosses. J. Anim. Sci. 23:105.
Sidwell, G. M. and L. R. Miller. 1971, Production in some pure breeds of sheep and their crosses. I. Reproductive efficiency in ewes. J. Anim. Sci. 32:1084. 54
Sidwell, G. M., D. 0. Everson and C. E. Terrill. 1962. Fertility, prolificacy and lamb livability of some pure breeds and their crosses. J. Anim. Sci. 21:875.
Terrill, C. E. and L. N. Hazel. 1947. Length of gestation in range sheep. Amer. J. Vet. Res. 8:66.
Thrift, F. A. and R. H. Dutt. 1972. Relationship between gestation length of artificially inseminated ewes and number, weight and sex of lambs born. J. Anim. Sci. 24:531.
Torrell, D. T., I. D. Hume and W. C. Weir. 1972. Effect of level of protein and energy during flushing on lambing performance of range ewes. J. Anim. Sci. 34:479.
Veseley, J. A. and H. F. Peters. 1972. Lamb growth performance of Romnelet, Columbia, Suffolk and N. C. Cheviot breeds and all single and three breed crosses among them. Can. J. Anim. Sci. 52:283.
Willham, R. L. 1969. Genetic consequence of crossbreeding. J. Anim. Sci. 30:690.
Wright, S. 1939. Genetic principles governing the rate of progress of livestock breeding. Proc. Amer. Soc. Anim. Prod., 32nd Ann. Meeting. 18-26. (A.B.A., 8:332). APPENDICES 55
APPENDIX A. BREED MEANS FOR FERTILITY
Overall 88.6%
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 85,7 91,9 91,8 89.8 Dam Suffolk 85.6 82.6 89,4 85.9 Breed
Willamette 90.8 91.8 87.8 90.2
Sire Breed 87.4 88.8 89.7 88.6 Averages
South Farm 87.9%
Hampshire Suffolk Willamette
Sire Breed 85.7 88.4 89.7
Dam Breed 88.4 88.4 87.1
Hill Pasture 89.3%
Hampshire Suffolk Willamette
Sire Breed 89.1 89.1 89.7
Dam Breed 91.2 83.3 93.3 56
APPENDIX B. BREED MEANS
Based on Ewes Mated
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages a Hampshire 1.39 1.44 1.35 1.39 b 1.15 1.21 1.19 1.19 97.3 110.1 100.9 102.8
Suffolk 1.54 1.33 1.50 1.46 1.31 1.79 1.31 1.27 Breed 120.8 109.6 124.5 118.3
Willamette 1.54 1.56 1.48 1.53 1,39 1.33 1.21 1.31 125.9 126.4 105.4 119.2
Sire Breed 1.49 1.45 1.44 1.46 Averages 1.29 1.24 1.24 1.26 114.6 115.4 110.3 113.4
a Born Total b Number Weaned c Pounds Weaned 57
Based on Ewes Lambing
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages a Hampshire 1.63 1.66 1.71 1.67 b 1.37 1.41 1.50 1.43 c 111.6 124.1 123.2 119.6 d 84.7 86.5 87.5 86.3 Dam Suffolk 1.66 1.58 1.61 1.62 Breed 1.39 1.38 1.39 1.39 127.6 126.0 130.2 127.9 85.7 90.1 89.0 88.3
Willamette 1.61 1.63 1.64 1.63 1.45 1.36 1.33 1.38 129.1 128.5 113,6 123.7 91.7 88.0 81.3 87.0
Sire Breed 1.64 1.62 1.65 1.64 Averages 1.41 1.38 1.40 1.40 122.8 126.2 122.3 123.8 87.4 88.2 85.9 87.2 a Born Total b Number Weaned dPounds Weaned Percent Weaned 58
APPENDIX C. BREED/YEAR/LQCATION MEANS
South, Farm Based Qn Ewes. Mated , Year 1
Sire Breed Dam Breed Hampshire Suffolk WillametteAverages
Hampshire 1.46 1.76 1.79 1.67 b 1.32 1.64 1.55 1.50 c 114.3 148.8 137.7 133.6 Dam Suffolk 2.04 1.74 2.15 1.98 Breed 1.77 1.56 2.02 1.78 170.5 140.4 186.5 165.8
Willamette 2.01 1.88 1.89 1.93 1.70 1.57 1.53 1.60 157.3 151.0 134.1 147.5
Sire Breed 1.84 1.79 1.94 1.84 Averages 1.60 1.59 1.70 1.63 147.4 146.7 152.8 149.0 a Born Total b Number Weaned c Pounds Weaned 59
South Farm - Based on Ewes Mated - Year 2
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.4% 1.15 1.22 1.26 1,00 .87 1.07 1.03 83.2 77.1 90.1 83.5
Dam Suffolk 1,34 1,16 1.19 1.23 1.00 .89 .90 .93 Breed 96.0 76,0 87.9 86.6
Willamette 1.16 1.34 .97 1.16 .95 ,98 .66 .86 85.5 83.1 62.8 77.1
Sire Breed 1.30 1.22 1.13 1.22 Averages .98 .91 .88 .92 88,2 78.7 80,3 82.4
South Farm - Based on Ewes Mated Year 3
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.04b 1.67 1.67 1.46 .61 1.39 1.34 1.11 36.4 105.1 96.3 79.3
Dam Suffolk 1.38 1.25 1.67 1.43 Breed 1.02 .95 1.31 1.09 82.2 72.7 105.2 86.7
Willamette 1.45 1.64 1.42 1.50 1.25 1.22 .95 1.14 93.0 101.3 69.8 88.0
Sire Breed 1.29 1.52 1.59 1.47 Averages .96 1.19 1.20 1.12 70.5 93.0 90.4 84,6
a Born Total b Number Weaned Pounds Weaned 60
Hill Pasture - Based on Ewes Mated - Year 1
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.59: 1.56 1.34 1.50 1.51 1.31 1.14 1.32 131.40 126.9 102.0 120.0
Dam / Suffolk 1.91 1.60 1.78 1.76 1.70 1.40 1.67 1.62 Breed 158.2 145.4 161.0 154.9
Willamette 1.98 1.74 1,90 1.87 1.73 1,55 1.64 1.64 163.8 158.9 142.5 155.1
Sire Breed 1.83 1.63 1.67 1.71 Averages 1.65 1.45 1.58 1.53 151.1 143,7 135.2 143.5 a Born Total b Number Weaned cPounds Weaned Hill Pasture - Based on Ewes Mated - Year 2
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.56a .98 .80 1.11 b 1.4 .75 .87 1.01 123.6 78.4 77.7 90.6 Dam Suffolk 1.24 1.05 .84 1.04 1.15 1.05 .77 .99 Breed 106.9 104.3 85.7 99.0
Willamette 1.15 1.22 1.12 1.16 1.20 1.17 1.08 1.15 115.2 114.2 100.6 110.0
Sire Breed 1.32 1.08 .92 1.10 Averages 1.25 .99 .91 1.05 115.2 96.3 88.0 99.9
Born Total b Number Weaned cPounds Weaned 61
Hill Pasture - Based on Ewes Mated- Year 3
Sire Breed Dam Breed Hampshire Suffolk WillametteAverages a Hampshire 1.26 1.55 1.30 1.37 1.06 1.31 1.18 1.18 94.6 124.2 101.7 106.8 Dam Suffolk Breed 1.32 1.19 1.38 1.30 1.21 1.16 1.22 1.20 110.9 118.8 120.8 116.8
Willamette 1.49 1.57 1.57 1.54 1.54 1.46 1.37 1.46 140.6 150.2 122.3 137.7
Sire Breed 1.36 1.44 1.42 1.41 Averages 1.27 1.31 1.26 1.28 115.4 131.1 114.9 120.5 a Born Total b Number Weaned c Pounds Weaned 62
South Farm - Based on Ewes Lambing - Year
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.74 1.91 1.95 1.86 b .57 1,78 1.69 1.68 123,0 149.8 136.2 136.3 d 91,3 94.4 84.1 90.0 Dam Suffolk 1.73 1,84 2,24 1.94 1.50 1.60 2.08 1.73 Breed 135.5 135.7 180.1 150.4 89.4 87.3 100.8 92.5
Willamette 2.00 1.90 1.93 1.94 1.65 1.54 1.93 1.56 140.7 139.5 120.6 133.6 84.9 83.4 74.9 81.1
Sire Breed 1.82 1.88 2.04 1.91 Averages 1.57 1.64 1.76 1.66 133.1 141.7 145.6 140.1 88.5 88.4 86.6 88.5
a Born Total b Number Weaned c Pounds Weaned d Percent Weaned 63
South Farm - Based on Ewes Lambing - Year 2
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.82143 1.70 1.65 1.72 1.32 1.28 1.41 1.34 104.0 110.3 115.6 110.0 d 70.1 74.8 83.2 76.0 Dam Suffolk 1.68 1.66 1.47 1.60 1.22 1.07 1.18 Breed 1.25 113.4 105,2 103.5 107.4 76.8 77,4 79.0 77.7
Willamette 1.55 1.66 1.43 1.55 1.26 1.21 1.01 1.16 110.4 103.4 91.8 101.9 84.8 80.3 76.8 80.6
Sire Breed 1.68 1.67 1.52 1.62 Averages 1.27 1.25 1.16 1.23 109.3 106.3 103.6 106.4 77.2 77.5 79.7 78.1 a Born Total b Number Weaned c Pounds Weaned d Percent Weaned 64
South Farm - Based on Ewes Lambing- Year 3
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.64a 1.84 1,79 1.76 b 1.08 1.59 1.49 1,39 77.34 111.4 104.3 d 67.1 90.9 87,2 81.7 Dam Suffolk 1.66 1.67 1.79 1.71 Breed 1.25 1.29 1.40 1.31 108.4 107.3 117.6 111.1 79.0 81.7 82.3 81.0
Willamette 1.77 1.81 1.67 1.75 1.53 1.34 1.15 1.34 123.8 115.7 89.7 109.7 87.9 82.9 71.6 80.8
Sire Breed 1.69 1.77 1.75 1.74 Averages 1.29 1.41 1.35 1.35 103.2 115.7 106.2 108.4 78.0 85.2 80.4 81.2 a Born Total b Number Weaned c Pounds Weaned d Percent Weaned 65
Hill Pasture - Based on Ewes Lambing- Year 1
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1,51a 1.56 1.74 1.60 1.50 1.33 1.45 1.43 120 9 4 119.3 117.9 119.4 100.2 85.9 81.2 89.1 D Suffolk 1.62 1.51 1.76 1.63 Breed 1.47 1.43 1.65 1.52 130.0 133,0 150.5 137.8 89.1 95.1 95.9 93,4
Willamette 1.64 1.54 1.77 1.65 1.42 1.54 1.50 1.43 126.5 135,0 122.4 128.0 88.5 86.9 79.7 85.0
Sire Breed 1.59 1.54 1.76 1.63 Averages 1.46 1.38 1.53 1.46 125.8 135.0 122.4 128.4 92.6 89.3 85.6 89.2 a Born Total b Number Weaned c Pounds Weaned d Percent Weaned 66
Hill Pasture - Based on Ewes Lambing- Year 2
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages
Hampshire 1.63a 1.39 1,49 1.50 1.42 1,08 1,43 1.33 128.5 106,5 124.0 119.7 d 92.7 80.0 94.1 88.9
Dam Suffolk 1.62 1.39 1.03 1.35 1.45 1.33 0.90 1.23 Breed 134.6 129.3 100.7 121.5 90.2 98.9 87.8 92.3
Willamette 1.24 1.35 1.42 1.34 1,28 1.30 1.34 1.31 122.9 125.6 123.2 123.9 102.1 97.5 96.7 98.8
Sire Breed 1.50 1.38 1.31 1.40 Averages 1.41 1.24 1.22 1.29 128,7 120.5 122.0 123.7 95.0 92.1 92.9 93.3
a Born Total b Number Weaned dPounds Weaned Percent Weaned 67
Hill Pasture - Based on Ewes Lambing Year 3
Sire Breed Dam Breed Hampshire Suffolk Willamette Averages a Hampshire 1.47 1.55 1.65 1.56 b 1.26 1.39 1.51 1,39 115.9 4 134.5 133.9 128.1 86.8 93.2 95.2 91.8
Suffolk 1.63 1.42 1.37 1.47 1.48 1,37 1.23 1.36 143.7 145.4 128.8 139.3 89.5 93.2 95,2 91.8
Willamette 1.48 1.52 1.63 1.54 1.56 1.42 1,44 1.47 150.4 152.0 133.8 145.4 102.3 97.3 87.9 95.8
Sire Breed 1.53 1.50 1.55 1.53 Averages 1.43 1.39 1.39 1.40 136.7 144.0 132.2 137.6 92.9 97.0 90.5 93.5 a Born Total b cNumber Weaned Pounds Weaned d Percent Weaned APPENDIX D. LOCATION MEANS
Based on Ewes Mated
Overall South Farm Hill Pasture
Fertility (%) 88,6 87.9 89.3
Born Total 1.46 1.51 1.41
Lambs Weaned 1.26 1.23 1.29
Pounds Weaned 113.4 105.3 121.5
Based on Ewes Which Lambed
Overall South Farm Hill Pasture
Born Total 1.64 1.76 1.52
Lambs Weaned 1.40 1,41 1,39
Pounds Weaned 123.8 118.3 129.3
Percent Weaned 87.2 82.4 92.0 69
APPENDIX E. HETEROSIS
Overall
Purebred Crossbred Percent Trait Mean Mean HeterOti
Fertility 85.9 90.2 5.0%
Per Ewe Mated
Born Total 1.40 1.49 6.4% Weaned 1,18 1.29 9,3% Pounds Weaned 104.1 118,1 13.4%
Per Ewe Lambing
Born Total 1,62 1;65 1.9% Number Weaned 1,36 4.4% Survivability 85,4 88.1 3,2% Pounds Weaned 117.1 127,1 8.5%
Year 1
Per Ewe Mated
Born Total 1.70 1.94 14.1% Weaned 1.47 1.71 16.3% Pounds Weaned 129.6 158.6 22.4%
Per Ewe Lambing
Born Total 1.84 1.96 6.5% Number Weaned 1.56 1.71 9.4% Survivability 84.5 89.5 5.9% Pounds Weaned 126.4 147.0 16.3%
Year 2
Per EWe'Mated
Born Total 1.18 1.23 4.5% Number Weaned .85 .96 12.9% Pounds Weaned 74.0 86.6 17.0% Purebred Crossbred Trait Mean Mean
Per Ewe Lambing
Born Total 1.64 Number Weaned 1,19 Survivability 74.8
Year 3
Per Ewe Mated
Born Total 1.24 1.58 27.4% Number Weaned .84 1.26 50.0% Pounds Weaned 59.6 97,2 63.1%
Per Ewe Lambing
Born Total 1,66 1,78 7.2% Number Weaned 1.17 1,43 22.5% Survivability 73.5 85.0 15.6% Pounds Weaned 91.4 116.9 27.8%
Hill Pasture
Year 1
Per Ewe Mated
Born Total 1.70 1.72 1.1% Number Weaned 1.55 1.52 -1.9% Pounds Weaned 139.8 145.1 3.8%
Per Ewe Lambing
Born Total 1.60 1.64 2.7% Number Weaned 1.48 1.45 -2.1% Survivability 91.7 87.9 -4.1% Pounds Weaned 125.4 130.0 3.6% 71
Purebred Crossbred Percent Trait Mean Mean Heterosis
Year 2
Per ...EWe.Mated
Born Total 1.24 1.04 -16.3% Number Weaned 1.18 .99 -16.1% Pounds Weaned 109,5 96.4 -12.0%
Per Ewe Lambing
Born Total 1.48 1.35 -8.6% Number Weaned 1.39 1.24 -10.8% Survivability' 96.1 92.0 -4.3% Pounds Weaned 127.0 119.1 -6,3%
Year 3
Per Ewe Mated
Born Total 1.34 1.44 7.1% Number Weaned 1.20 1,32 10.0% Pounds Weaned 111.9 124.7 11.4%
Per Ewe Lambing
Born Total 1.51 1.53 1.5% Number Weaned 1.36 1.43 5.3% Survivability 91.7 94.3 2,8% Pounds Weaned 131.7 140.6 6.7%